A real-time quasi-3D metal flow model for hot strip rolling. (August 2019)

• Record Type: Journal Article
• Authors: Yao, Chihuan; He, Anrui; Shao, Jian; Zhao, Jianwei
• Is Part Of: International journal of mechanical sciences. Volume 159(2019)
• Page Start: 91
• Page End: 102
• Publication Date: 2019-08
• Subjects: Strip deformation -- Lateral spread -- Shape prediction -- Finite difference method -- Finite element analysis; Mechanical engineering -- Periodicals; Génie mécanique -- Périodiques; Mechanical engineering; Maschinenbau; Mechanik; Zeitschrift; Periodicals; 621.05
• DOI: 10.1016/j.ijmecsci.2019.05.045
• Languages: English
• ISSNs: 0020-7403

Abstract

Highlights: A novel metal flow model for shape control in hot strip rolling is proposed. The model predicts the distributions of lateral flow, tension and rolling force. The calculation time is less than 0.1 s, capable of real-time application. A linearization process ensures the efficient and robust solution. The lateral flow is found important for the shape prediction in hot strip rolling. Abstract: A new model is presented for predicting the distributions of lateral metal flow, tension and rolling pressure in hot strip rolling, with a fast speed capable of real-time application in the shape control system, helping to achieve high-quality flat products and rolling stability. To reduce the computational cost, a quasi-3D approximation is performed, preserving the essential physics of the problem. The equations of rolling deformation are decoupled, linearized and then discretized by the finite difference method to establish a global system of linear equations and achieve efficiency and robustness of the iterative solution. The model is validated by comparisons of the predictions with those obtained by the full-scale finite element simulations. In all thick and thin cases with various exit strip crowns, the results match, with errors of less than 15%, while reducing the calculation time to 0.1 s. It is found that the lateral flow is highly coupled with tension and rolling pressure, and is therefore important for the shape prediction in hot strip rolling. The metal flow model is also complemented with a roll stack deformation model to predict strip profile, and comparisons with industrial experiment measurements show a satisfactory accuracy.